Cetuximab with chemotherapy as 1st-line treatment for metastatic colorectal cancer: a meta-analysis of the CRYSTAL and OPUS studies according to KRAS and BRAF mutation status *C.H. Köhne, P. Rougier, C. Stroh, M. Schlichting, C. Bokemeyer, E. Van Cutsem * Klinikum Oldenburg, Oldenburg, Germany

Background The randomized, controlled CRYSTAL1 and OPUS2 trials showed that adding cetuximab, to standard 1st-line irinotecan- or oxaliplatin-based therapy improved efficacy in patients with KRAS wild-type (wt) mCRC The serine-threonine kinase BRAF is a direct downstream effector of KRAS BRAF gene mutations have been detected in 8% of CRC tumors3 BRAF mutation status has been suggested to be predictive of cetuximab efficacy in pretreated patients with mCRC4 1Van Cutsem E, et al. N Engl J Med 2009;360:1408-17 2Bokemeyer C, et al. J Clin Oncol 2009;27:663-71 3Roth A, et al. J Clin Oncol 2009 E pub doi/10.1200/JCO.2009.23.3452 4Di Nicolantonio F, et al. J Clin Oncol 2008;26:5705-12

Updated: Pooled analysis of the CRYSTAL and OPUS studies The effect of tumor KRAS and BRAF mutation status on PFS best overall response (OR) OS in patients receiving cetuximab plus standard 1st-line chemotherapy was investigated further in this updated pooled analysis of individual patient data from the CRYSTAL and OPUS populations presented at ECCO 20091 The ascertainment rate for KRAS mutation status in the CRYSTAL and OPUS studies has been enhanced from the published studies and the most recent survival data have been used for analysis of efficacy endpoints and tumor KRAS and BRAF mutation status 1Van Cutsem E, et al Eur J Cancer Supplements 2009;7:S345, P-6077

Pooled analysis: Study objectives To investigate the effects of tumor KRAS and BRAF mutation status on the efficacy of cetuximab in patients with KRAS wt tumors Using the key efficacy endpoints from the two trials: PFS (CRYSTAL) OR (OPUS) And the secondary endpoints as defined in the trial protocols

Patients and samples Patients were randomized to receive cetuximab in combination with FOLFIRI (CRYSTAL)1 or FOLFOX4 (OPUS)2 or the standard 1st -line treatment alone The number of samples evaluable for KRAS mutations Increased from 540/1198 (45%)1 to 1063 (89%) in the CRYSTAL study Increased from 233/337 (69%)2 to 315 (93%) in the OPUS study The number of samples evaluable for BRAF mutations Increased from 529/1198 (44%) previously reported6 to 1000 (83%) in the CRYSTAL study Were 309 (91%) in the OPUS study In the CRYSTAL and OPUS studies 625 and 175 KRAS wt tumors respectively, were evaluable for BRAF mutation status 1Van Cutsem E, et al. N Engl J Med 2009;360:1408-17 2Bokemeyer C, et al. J Clin Oncol 2009;27:663-71 3 Köhne C, et al J Clin Oncol 2009;27:15s (suppl;abstr 4068)

CRYSTAL and OPUS: Assessments For each trial: Primary analyses of PFS and OR rate were based on CT or MRI scans as assessed by an independent radiology review committee according to modified WHO criteria Additional tumor mutation analysis was performed on material extracted from stained slides previously collected to evaluate tumor EGFR expression status KRAS (codons 12/13) and BRAF (V600E) mutations were detected using a polymerase chain reaction clamping and melting curve technique

Statistical considerations The pooled analysis was performed on individual patient data from the two trials CRYSTAL study data cuts-offs were: PFS (IRC), OR (IRC) on the 27th July 2006, OS on the 31st May 2009 OPUS study data cut-offs were: PFS (IRC), OR (IRC) on the 1st March 2007, OS on the 30th November 2008 Treatment groups were compared in a Cox proportional hazards model for OS and PFS and in a logistic regression model for OR that were both adjusted for study effects and stratified by ECOG PS Inter-study heterogeneity was assessed by testing for individual treatment effect estimates across the studies

Baseline characteristics of patients by KRAS and BRAF mutation status KRAS wt (n=845) KRAS wt/BRAF wt (n=730) KRAS wt/BRAF mt (n=70) Characteristics FOLFOX4 (n=447) Cetuximab +FOLFOX4 (n=398) (n=381) Cetuximab +FOLFOX4 (n=349) (n=38) Cetuximab +FOLFOX4 (n=32) Gender, n (%) Male 266 (60) 238 (60) 228 (60) 214 (61) 22 (58) 17 (53) Female 181 (40) 160 (40) 153 (40) 135 (39) 16 (42) 15 (47) Median age, yrs (range) 59 (19–84) 61 (24–79) (25–75) 64 (34–79) Age categories, n (%) <65 yrs 297 (66) 246 (62) 254 (67) 216 (62) 25 (66) 18 (56) ≥65 yrs 150 (34) 152 (38) 127 (33) 133 (38) 13 (34) 14 (44) ECOG PS 0-1, n (%) 423 (95) 379 (95) 363 (95) 332 (95) 34 (90) 32 (100) Liver-metastasis only, n (%) 95 (21) 93 (23) 85 (22) 80 (23) 4 (11) 10 (31) Involved disease sites ≤2, n (%) 370 (83) 344 (86) 321 (84) 305 (87) 25 (78) Prior adjuvant chemotherapy , n( %) 90 (20) 85 (21) 83 (22) 87 (25) 6 (16) 5 (16) Cancer Com: Due to space constraints only % values have been shown for most patient demographics, the numbers are shown in the poster. ECOG PS, Eastern Cooperative Group performance status; mt, mutant wt, wild-type; yrs, years

Clinical efficacy by KRAS/BRAF tumor mutation status KRAS wt (n=845) KRAS wt/BRAF wt (n=730) KRAS wt/BRAF mt (n=70) FOLFIRI   (n= 447) Cetuximab + FOLFIRI (n= 398) (n= 381) Cetuximab +FOLFIRI (n= 349) (n=38) (n=32) Median OS months [95% CI] 19.5 [17.8–21.1] 23.5 [20.7–25.7] 21.1 [19.5–23.6] 24.8 [22.1–27.0] 9.9 [5.7–13.6] 14.1 [8.8–18.5] Hazard ratioa p-valueb (heterogeneity p-value)c 0.81 [0.69–0.94] 0.0062 (0.6996) 0.84 [0.71–1.00] 0.0479 (0.6980) 0.62 [0.36–1.06] 0.0764 (0.0478) Median PFS months 7.6 [7.4–8.4] 9.6 [8.9–11.3] 7.7 [7.4–9.0] 10.9 [9.2–11.9] 3.7 [2.1–7.9] 7.1 [3.7–9.1] 0.66 [0.55–0.80] <0.0001 (0.3332) 0.64 [0.52–0.79] (0.3362) 0.67 [0.34–1.29] 0.2301 (0.3778) OR rate (%) 38.5 57.3 40.9 60.7 13.2 21.9 Odds ratioa 2.16 [1.64­–2.86] (0.5568) 2.27 [1.68–3.07] (0.5891) 1.60 [0.45–5.67] 0.4606 (0.1727) Would you like this table split into two parts. Part I KRAS wt Part II KRASwt/BRAFwt, KRAS wt BRAF mt aHazard ratios <1 for OS and PFS and odds ratios >1 for OR indicate a benefit for the addition of cetuximab to chemotherapy compared with chemotherapy alone; bLikelihood ratio test on treatment effect in stratified Cox’s proportional hazards model;cPooled (stratified) likelihood ratio test on study treatment interaction in stratified Cox’s proportional hazards model CI, confidence interval; OR, best overall response; OS, overall survival; PFS, progression-free survival; mt, mutant; wt, wild-type 9

Pooled analysis of OS in patients with KRAS wt tumors Study KRAS wt population Hazard Ratio [95% CI] Pooled analysis n=845 0.81 [0.69-0.94] 0.80 [0.67-0.95] CRYSTAL n=666 0.85 [0.60-1.22] OPUS n=179 1.0 1.3 2.0 0.7 0.5 Benefit under cetuximab + chemotherapy Benefit under chemotherapy

Pooled analysis of PFS in patients with KRAS wt tumors Study KRAS wt population Hazard Ratio [95% CI] 1.0 1.3 2.0 0.7 0.5 Pooled analysis n=845 0.66 [0.55-0.80] CRYSTAL n=666 0.70 [0.56-0.87] OPUS n=179 0.57 [0.38-0.86] Benefit under cetuximab + chemotherapy Benefit under chemotherapy

Pooled analysis of OR in patients with KRAS wt tumors Study KRAS wt population Odds Ratio [95% CI] 1.0 1.3 2.0 0.7 0.5 Pooled analysis n=845 2.16 [1.64-2.86] CRYSTAL n=666 2.07 [1.52-2.83] OPUS n=179 2.55 [1.38-4.72] Benefit under chemotherapy Benefit under cetuximab + chemotherapy

Probability of overall survival Pooled analysis of OS by treatment group for patients with KRAS wt tumors 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 FOLFIRI / FOLFOX4 Cetuximab + FOLFIRI / FOLFOX4 (n=447) (n=398) No. of events 359 297 Median OS 19.5 months 23.5 months [95% CI] [17.8-21.1] [20.7-25.7] HR [95% Cl] 0.81 [0.69-0.94] p-value 0.0062 Probability of overall survival Cetuximab + FOLFIRI / FOLFOX4 FOLFIRI / FOLFOX4 18 6 12 24 60 30 36 42 48 54 Time (months) Number of patients Cetuximab + CT 398 356 296 246 177 128 83 65 21 4 CT 447 395 313 227 159 112 67 48 18 2 CI, confidence interval; CT, chemotherapy; PFS, progression-free survival

Probability of progression-free survival Pooled analysis of PFS by treatment group for patients with KRAS wt tumors 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 FOLFIRI / FOLFOX4 Cetuximab + FOLFIRI / FOLFOX4 (n=447) (n=398) No. of events 251 184 Median PFS 7.6 months 9.6 months [95% CI] [7.4-8.4] [8.9-11.3] HR [95% Cl] 0.66 [0.55-0.80] p-value <0.0001 Probability of progression-free survival Cetuximab + FOLFIRI / FOLFOX4 FOLFIRI / FOLFOX4 4 8 12 16 20 Time (months) Number of patients Cetuximab + CT 398 286 154 46 9 1 CT 447 298 128 24 4 CI, confidence interval; CT, chemotherapy; PFS, progression-free survival

Sensitivity analysis Results of sensitivity analyses for fixed or random effect models were nearly identical thus confirming the robustness of the data

Conclusions In this pooled analysis, a significant improvement in OS time was demonstrated for patients with KRAS wt tumors receiving cetuximab plus chemotherapy compared with chemotherapy alone This analysis strengthens the findings from the CRYSTAL and OPUS studies that the addition of cetuximab to 1st-line chemotherapy in patients with KRAS wt mCRC significantly improves OR and PFS The cetuximab treatment effect does not vary by BRAF mutation status. Patients with BRAF mutations also appear to benefit from cetuximab The pooled analysis confirms the consistency of the benefit across all endpoints obtained with cetuximab added to 1st-line chemotherapy in patients with KRAS wt mCRC